My own point of view, on the contrary, has been general and reproductive, for my classification is based upon the natural causes of crime, individual, physical, and social, and to this extent it corresponds more closely with the theoretical and practical requirements of criminal sociology. If the curative art of society, like that of individuals, expects from positive knowledge an indication of remedies, it is clear that a classification based on the fundamental causes of crime is best fitted to indicate a social cure for this manifestation of disease, which is the essential object of criminal sociology. For, as in biology one is carried from purely descriptive anatomy to genetic anatomy and physiology, so in sociology we must pass on from purely legal descriptions of crimes to the genetic knowledge of the criminals who commit these crimes. For this reason all the chief classifications of criminals, as has been seen, may be brought into <p 50>line with my own, by virtue of the more complete and fruitful test which has established it. And thus we have a manifest proof that this classification actually represents the common and permanent basis of all the chief anthropological categories of criminals, whether in regard to their natural causality and their specific character, or in regard to the different forms of social self-defence which spring out of them, and which must be adapted to the natural causes of crime, and to the principal criminal types. But whatever classification may be accepted, we shall always have, as the fundamental axiom of criminal anthropology, this variety in the types of criminals, which must henceforth be indispensable to all who are theoretically or practically concerned with crime. CHAPTER II. _THE DATA OF CRIMINAL STATISTICS._ FOR moral and social facts, unlike physical and biological facts, experiment is very difficult, and frequently even impossible; observation in this domain brings the greatest aid to scientific research. And statistics are amongst the most efficacious instruments of such observation. It is natural, therefore, that criminal sociology, after studying the individual aspect of the natural genesis of crime, should have recourse to criminal statistics for the study of the social aspect. Statistical information in the words of Krohne, ``is the first condition of success in opposing the armies of crime, for it discharges the same function as the Intelligence department in war.'' From statistics, in fact, the modern idea of the close relation between offences and the conditions of social life, in some of its aspects, and above all in certain particular forms, has most directly sprung.